{{ normalized.worksheet_title }}
{{ normalized.student_instructions }}
Word bank
{{ segment.text }}{{ segment.text }}
{{ normalized.student_instructions }}
{{ segment.text }}{{ segment.text }}
| Blank | Answer | Selection | Word # | Context | Copy |
|---|---|---|---|---|---|
| {{ row.number }} | {{ row.answer }} | {{ row.reason_label }} | {{ row.word_number }} | {{ row.context }} |
{{ check.note }}
A cloze activity removes selected words or phrases from a passage and asks the reader to restore them from context. The surrounding grammar and meaning remain visible, so a blank can test vocabulary, sentence structure, topic knowledge, or careful reading depending on which text is removed.
Blank selection determines what the worksheet measures. Removing every fifth word samples the whole passage mechanically. Removing a vocabulary list concentrates practice on chosen terms. Removing the most frequent content words emphasizes repetition in the passage, while bracketed answers let a teacher force an exact word or phrase regardless of the automatic mode.
Difficulty depends on more than density. A rare technical term is harder than a familiar connector, a multiword phrase removes more context than one short word, and answer-length blanks can reveal more than fixed-width lines. Repeated answers may make a bank easier because one choice fits several places, while removing attached punctuation may eliminate useful grammar clues.
Passage quality still sets the ceiling. A worksheet cannot repair an unclear, inaccurate, inaccessible, or poorly leveled source. Read the student version without the key and check that each intended answer is recoverable from the text or the stated support.
Cloze work shows how a learner uses the available clues; it does not by itself prove full comprehension of the passage. Follow it with explanation, discussion, or another task when understanding beyond word restoration matters.
Begin with a clean passage and decide whether the blanks should sample the text or target specific learning.
Use blank density as a review cue, not a grade-level standard. The generated teacher check labels density below 8% as light, 8% through 22% as balanced, above 22% through 35% as demanding, and above 35% as very dense.
The passage is tokenized into words, numbers, bracketed phrases, and unchanged text between them. Apostrophes and hyphens may remain inside a word. A bracketed phrase counts the words it contains but becomes one blank and one answer-ledger row.
Blank density measures how much of the tokenized passage has been removed.
B is the number of generated blanks and W is the passage word count. A phrase contributes its number of words to W but only one to B. The displayed check rounds density to one decimal place; selection uses the unrounded value, and the very-dense warning is triggered only when D is greater than 35%.
Bracketed answers are selected first. Remaining tokens follow the chosen automatic rule, and the maximum-blank cap is applied in passage order after selection.
| Mode | Selection rule | Important boundary |
|---|---|---|
| Every nth word | After the kept lead words, select positions whose offset is divisible by the interval. | Interval 2 to 50; lead words 0 to 50. |
| Vocabulary list | Match single terms and longest multiword phrases; phrases are considered before single words. | Matching is case-insensitive unless case sensitivity is enabled. |
| Most frequent words | Exclude built-in and added stop words and terms below the minimum length, then rank by frequency, first occurrence, and alphabetical order. | Select 1 to 30 distinct terms; each occurrence of a selected term becomes eligible. |
| Manual only | Select bracketed words or phrases and leave all other text visible. | At least one bracketed answer is needed to produce a worksheet. |
For nth-word mode, a non-bracketed word at position n is selected when it comes after the kept lead words L and its offset is an exact multiple of interval I.
Fixed-width blanks use 4 to 30 underscores. Answer-length blanks use the answer’s character count, with a minimum of 4 and a maximum of 30. Numbering is optional. Removing punctuation strips attached opening and closing marks around selected answers; it does not rewrite punctuation elsewhere.
The bank begins with answers in passage order. Duplicate removal compares normalized answers under the selected case rule. A shuffled bank uses the seed to produce the same order from the same answers, while recall mode emits no bank. A maximum of 0 means uncapped; otherwise the accepted cap is 1 through 300.
The passage accepts up to 50,000 characters. Vocabulary targets accept 4,000 characters, added stop words 2,000, the seed 80, the title 90, and student instructions 180. Validation stops when the passage is empty, has no recognized words, or the chosen rules create no blanks.
The passage and TXT file are read in the browser and are not sent to a server by this worksheet generator. A TXT file must be plain text and no larger than 1 MB.
Automatic selection cannot judge whether a blank is instructionally appropriate, ambiguous, or recoverable for a particular learner. Review the student copy, bank, answer key, phrase handling, repeated answers, and attached punctuation before printing or sharing it.
For “One two three four five six” with interval 3 and no kept lead words, the third and sixth words become blanks. That produces 2 blanks from 6 words, or 33.3% density. An ordered bank contains “three” followed by “six.”
In “The water cycle repeats and condensation forms,” targets water cycle and condensation create two blanks. The two-word target stays together as one blank and one ledger entry, while its two words still count toward the seven-word density denominator.